[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Libcdio-devel] Re: Submission of new mmc function for libcdio
From: |
Thomas Schmitt |
Subject: |
[Libcdio-devel] Re: Submission of new mmc function for libcdio |
Date: |
Fri, 29 Jan 2010 08:24:33 +0100 |
Hi,
> There should be an additional parameter
> which is basically a driver_return_code_t that indicates whether the answer
> is valid or not, and if not gives some idea of what went wrong.
Good idea.
> The caller
> of this can also query SCSI sense data if this "valid" parameter says
> something went wrong.
There is a little problem with obtaining sense
reply in higher function levels:
The caller of mmc_get_disc_erasable() would have
to make the assumption that the function
performed only a single MMC command resp. that
the failed command was the last command
performed in that function.
This would hamper to implement complex MMC
gestures inside mmc_get_disc_erasable().
So if the sense reply is of meaning for
callers of callers of mmc_run_cmd(), then the
immediate caller of mmc_run_cmd() has to take
care of the sense reply.
Further: the sense is really meaningful only if
one knows from what MMC command it stems.
So i advise to convert the sense reply into a
libcdio error reply already inside the function
that directly calls mmc_run_cmd().
I admit mmc_get_disc_erasable() is not much of
a candidate for complexity. But in the general
case one should try to achieve good encapsulation
of the function entrails.
Have a nice day :)
Thomas